how the mass media divide us - university of pennsylvania

University of Pennsylvania University of Pennsylvania

ScholarlyCommons ScholarlyCommons

Departmental Papers (ASC) Annenberg School for Communication

1-1-2006

How the Mass Media Divide Us How the Mass Media Divide Us

Diana C. Mutz University of Pennsylvania, [email protected]

Follow this and additional works at: https://repository.upenn.edu/asc_papers

Part of the Mass Communication Commons

Recommended Citation (OVERRIDE) Recommended Citation (OVERRIDE) Mutz, D. C. (2006). How the mass media divide us. In D. Brady & P. Divola (Eds.), Red and blue nation (pp. 223-248). Washington DC: Brookings Institution Press. Retrieved from http://repository.upenn.edu/asc_papers/126

This paper is posted at ScholarlyCommons. https://repository.upenn.edu/asc_papers/126 For more information, please contact [email protected].

https://repository.upenn.edu/

https://repository.upenn.edu/asc_papers

https://repository.upenn.edu/asc

https://repository.upenn.edu/asc_papers?utm_source=repository.upenn.edu%2Fasc_papers%2F126&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/334?utm_source=repository.upenn.edu%2Fasc_papers%2F126&utm_medium=PDF&utm_campaign=PDFCoverPages

https://repository.upenn.edu/asc_papers/126

mailto:[email protected]

How the Mass Media Divide Us How the Mass Media Divide Us

Abstract Abstract The chapters in this book suggest that scholars are nowhere near a consensus on whether the mass public is more polarized than it has been in the past and, if it is, relative to precisely when. Nonetheless, among those who believe the mass public has, indeed, become increasingly polarized in its views, mass media are very likely to be invoked as a cause. Perhaps this should come as no surprise - throughout American history, mass media have been blamed for just about every social ill that has befallen the country.

But in the midst of so much disagreement about when and whether and among whom this phenomenon has occurred, why is there so much agreement that media must somehow be to blame? A consensus on this point exists not so much because the empirical evidence is overwhelming, but because there are multiple theories that predict and explain how media might logically influence levels of mass polarization. Furthermore, it is possible to view mass media as engines of polarization even if one believes the public in general has not become polarized to any significant degree. Mass media are, after all, only one influence in a much larger system of institutions and influences.

For purposes of this chapter, I set aside the question of whether and to what extent mass polarization has occurred, in favor of an exploration of ways in which media have been implicated in polarizing processes. How might mass media be contributing to this widely decried state of affairs? And even if the public has not polarized, how might mass media nonetheless be encouraging mass opinion in more extreme directions?

Disciplines Disciplines Mass Communication

This book chapter is available at ScholarlyCommons: https://repository.upenn.edu/asc_papers/126

https://repository.upenn.edu/asc_papers/126

red and bluenation?

Copyright 2006, the Brookings Institution and the Board of Trustees of Leland Stanford Junior University.

red and bluenation?

hoover institution on war, revolution and peace

Stanford University

brookings institution pressWashington, D.C.

volume one

Characteristics and Causes of America’s Polarized Politics

pietro s. nivoladavid w. brady

editors


Copyright © 2006the brookings institutionboard of trustees of leland stanford junior university

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means without permission in writing from the Brookings Institution Press, 1775 Massachusetts Avenue, N.W., Washington, D.C. 20036, www.brookings.edu.

Red and Blue Nation? Characteristics and Causes of America’s Polarized Politics may be ordered from:brookings institution pressc/o HFS, P.O. Box 50370, Baltimore, MD 21211-4370Tel: 800/537-5487; 410/516-6956; Fax: 410/516-6998

Library of Congress Cataloging-in-Publication dataRed and blue nation? : characteristics and causes of America’s polarized politics /Pietro S. Nivola and David W. Brady, editors.

p. cm.Summary: “Considers the extent to which polarized views among political leaders and activists are reflected in the population at large. Pays particular attention to factors such as the increased influence of religion and the changing nature of the media and offersthoughtful analyses of the underlying problems”—Provided by publisher.Includes bibliographical references and index.ISBN-13: 978-0-8157-6082-5 (cloth : alk. paper)ISBN-10: 0-8157-6082-5 (cloth : alk. paper)ISBN-13: 978-0-8157-6083-2 (pbk. : alk. paper)ISBN-10: 0-8157-6083-3 (pbk. : alk. paper)

1. Political parties—United States. 2. Party affiliation—United States. 3. Polarization (Social sciences) 4. United States—Politics and government. I. Nivola, Pietro S. II. Brady, David W. III. Title.JK2261.R28 2006324.273—dc22 2006034595

9 8 7 6 5 4 3 2 1

The paper used in this publication meets minimum requirements of the American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials: ANSI Z39.48-1992.

Typeset in Adobe Garamond

Composition by Circle GraphicsColumbia, Maryland

Printed by R. R. DonnelleyHarrisonburg, Virginia


The chapters in this book suggest that scholars are nowhere near a consensuson whether the mass public is more polarized than it has been in the past

and, if it is, relative to precisely when. Nonetheless, among those who believethe mass public has, indeed, become increasingly polarized in its views, massmedia are very likely to be invoked as a cause. Perhaps this should come as nosurprise—throughout American history, mass media have been blamed for justabout every social ill that has befallen the country.

But in the midst of so much disagreement about when and whether andamong whom this phenomenon has occurred, why is there so much agreementthat media must somehow be to blame? A consensus on this point exists not somuch because the empirical evidence is overwhelming, but because there aremultiple theories that predict and explain how media might logically influencelevels of mass polarization. Furthermore, it is possible to view mass media asengines of polarization even if one believes the public in general has not becomepolarized to any significant degree. Mass media are, after all, only one influencein a much larger system of institutions and influences.

For purposes of this chapter, I set aside the question of whether and to whatextent mass polarization has occurred, in favor of an exploration of ways inwhich media have been implicated in polarizing processes. How might mass

How the Mass Media Divide UsDiana C. Mutz

5

223Copyright 2006, the Brookings Institution and the Board of Trustees of Leland Stanford Junior University.

media be contributing to this widely decried state of affairs? And even if thepublic has not polarized, how might mass media nonetheless be encouragingmass opinion in more extreme directions?

More from Which to Choose

One obvious and unmistakable characteristic of the current media landscape ischoice. No one questions the idea that there is more choice, that there are moremedia outlets now than thirty or forty years ago. As political scientist MarkusPrior notes in a forthcoming book, television provided a mere seven channels to theaverage household in 1970, and the three broadcast networks captured 80 percentof all viewing. By 2005, over 85 percent of households had cable or satellite access,and the average viewer had a choice of about a hundred channels. There arenow more broadcast networks than in 1970, but they collectively capture only40 percent of viewing.1 When one adds to this the choices offered by “newmedia”—the Internet, cell phones, iPods, and the like—then it seems indisputablethat Americans have more media choice than ever before.

There are two different theories that tie increased media choice to higher levelsof political polarization. One argument points to the increased number of choicespeople have in sources of political news as a cause of polarization. An altogetherdifferent theory suggests that it is the expansion of choices of nonpolitical enter-tainment media content that is most consequential for mass polarization.

Choosing Sources of Political News

Greater choice in sources of political news is relevant to political polarizationbecause it means that people must decide on some basis which sources to useand which not to use. More optimistic scholars have suggested that perhaps thiswill lead the United States toward a better approximation of a true “marketplace”of ideas, one in which people are exposed to many different perspectives and canweigh a broader range of views in formulating their opinions than was once thecase. This idealistic vision is appealing, but for the most part scholars havefocused on the other side of this double-edged sword: more choice also means thatpeople can more easily limit their exposure based on their own predispositions.

To the extent that the many sources of political news have identifiable politicalcomplexions, some people may end up choosing news sources that reinforce andintensify their preexisting views. In academic jargon, this phenomenon goes by

224 diana c. mutz

1. Prior (forthcoming 2007).


the name of “selectivity” or “biased assimilation of information.” Selectivity cantake place at several junctures with respect to mass media, including exposure toa particular source of political news, attention to what the source says, and biasedinterpretation when processing the content of political news. Because exposureto a source of political news is a prerequisite for any of the subsequent kinds ofselection to become relevant, selective exposure has been the target of the greatestresearch attention.

Concerns about selective exposure date back to the very earliest selection studiesconducted in the 1940s by Paul Lazarsfeld and his colleagues at ColumbiaUniversity.2 These early studies became closely linked with psychologist LeonFestinger’s theory of cognitive dissonance.3 Festinger suggested that people wantto avoid information that conflicts with their preexisting beliefs, and that theyseek out information—through activities such as selective exposure—that con-firms their current beliefs. Although a trickle of studies of selective exposure topartisan information continued over the past five decades, a renewed interest inthe topic has been spawned by the relatively recent proliferation of news outlets.Moreover, as a result of television programs that now appear more unabashedlypartisan, more scholars are finding this hypothesis and its potential polarizinginfluence worthy of study.

What do these recent studies suggest about whether selective exposure tolike-minded political content occurs and thus contributes to a more extremepublic? Do people select media content that is compatible with their own views andavoid exposure to alternative opinions? There is, for better or worse, no simpleanswer to these questions, in part because the internal context of these studiesvaries, but also because the external context—that is, the media environment—haschanged so tremendously since the 1940s, when the hypothesis was formulated.

Studies of the real-world context in which media choices are made tend tosupport the selective exposure hypothesis, but with important limitations. If onecompares the political views promoted by any given medium with the politicalviews held by that medium’s audience, investigators regularly find significantrelationships. For example, conservatives are more likely to read conservativenewspapers, and liberals are more likely to read liberal ones.

Unfortunately, such correlations are very limited in what they tell us aboutthe public’s propensity to selectively expose itself to political information. Mostobviously, it is possible that the partisan leanings of the media sources influenced

how the mass media divide us 225

2. Lazarsfeld, Berelson, and Gaudet (1948).3. Festinger (1957).


the readers, viewers, or listeners, so that they ended up holding similar politicalperspectives. In this case, a relationship between the partisanship of a newssource and the partisanship of its audience cannot necessarily be attributed toselective exposure. For example, in David Barker’s panel study of Rush Limbaughlisteners, he finds that regular listeners developed an antipathy toward RushLimbaugh’s favorite targets, thus suggesting that influence as well as selectiveexposure was occurring.4

Yet another prominent reason that these kinds of relationships are often dis-missed is because of the lack of direct evidence that people are actually makingchoices motivated by a desire to avoid dissimilar perspectives. In their influentialreview of the literature on both political and nonpolitical forms of selectiveexposure, David O. Sears and Jonathan L. Freedman suggested that the similar-ity between audiences’ opinions and the opinions of their media sources wasbecause of “de facto selective exposure.”5 According to this line of thought, acorrelation exists between the opinions of an information source and its audi-ence, but not because the audience is motivated to avoid disagreeable views, orbecause it is influenced by the media. Instead, the correlation arises because mediaenvironments tend to supply more like-minded sources of information to thebulk of the populations they reach than they supply dissonant sources. Red statestend to have more conservative newspapers than blue states, but according tothe de facto interpretation it is not because Republicans have chosen not to readmore liberal newspapers—thus driving them out of business. Instead, they readconservative newspapers because, in conservative regions, they are more widelyavailable than liberal ones.

Of course, a consideration of market forces turns this argument into a bit ofa chicken-and-egg debate. For instance, does Provo, Utah, not have any liberal-leaning media because no one there would choose to consume it, or can no onechoose it because it is not available there? In their review of the evidence in the1960s, Sears and Freedman emphasized the latter, suggesting that much of whathad been deemed as evidence that audiences were motivated to avoid oppositionalpolitical perspectives was really evidence of de facto selective exposure.

However, the preponderance of evidence cited in recent studies confirms thatpartisan audiences do select media that lean in the direction of their own views.And the evidence of such a relationship is common across virtually all forms ofmedia. Talk-radio listeners are likely to have views similar to those of the host

226 diana c. mutz

4. Barker (2002).5. Sears and Freedman (1967).


whose program they listen to, Republicans are more likely to be viewers of theright-leaning Fox News Channel, and political films such as Fahrenheit 9/11 areattended primarily by those who are already sympathetic to the films’ politicalperspectives. Likewise, visitors to a candidate’s website will predominantly alreadybe the candidate’s supporters.6 Even when political differences in media are fairlymuted, such as when there are two competing daily newspapers in the same city,there is evidence that people systematically choose to read the newspaper withan editorial slant closer to their own politics.7 In this case, availability is heldconstant (making the de facto interpretation moot), providing clear evidence ofpolitically motivated selective exposure.8

Although it seems logical to conclude from this that partisans on both sideswould, as a result, polarize further in the direction of their original views, thisconsequence is not yet well documented. Studies of small group interactionshave suggested that like-minded company leads to greater attitude extremity,but there is little evidence to date that documents that same process of influencewith like-minded mass media.9 Moreover, if this process of reinforcement hasbeen taking place all along, why have scholars only begun to suspect selectiveexposure as a cause of greater polarization in recent years?

Heightened concern about mass media’s role in polarization stems from anincrease in partisan differences across the many news sources now available tocitizens. Selective exposure requires that people associate particular news sourceswith particular brands of politics. So although it is unlikely that people todayare any more selective than they ever were, the media environment itself hasmade being selective easier than it was in the past, just by offering more diversesources. Just how much easier selectivity is now remains unclear. Few mainstreamnews sources bill themselves explicitly as Republican or Democratic in orientation,though increasingly some use terms such as “progressive” and “traditional” to cuetheir audiences as to what they can expect.

The way in which openly partisan media facilitate selective exposure is abun-dantly clear when one compares audience patterns in the United States with


6. See Cappella, Turow, and Jamieson (1996); Annenberg Public Policy Center, “Fahrenheit9/11 Viewers and Limbaugh Listeners about Equal in Size Even Though They Perceive Two Dif-ferent Nations, Annenberg Data Show,” August 3, 2004; Bimber and Davis (2003).

7. Mutz and Martin (2001).8. Ironically, having a choice of daily newspapers means that readers will be exposed to fewer

oppositional viewpoints since few people bother to read multiple papers. More voices means morechoice, but having choice means that people will pick and choose content that comports with theirown viewpoints and thus makes it less likely that they will hear the other side.

9. For a review of the evidence of small group interactions, see Sunstein (2002).


those in a country such as Great Britain, where newspapers openly carry parti-san labels. In Great Britain, newspaper readers perceive political agreement intheir media in roughly the U-shaped pattern shown in figure 5-1. The pattern ofnewspaper readership in Great Britain demonstrates classic evidence of partisan-based selective exposure (with Conservative and Labour parties replacing Repub-lican and Democrat, respectively). Strong partisans are the most motivated toselectively expose, but weaker partisans also do so to a lesser extent. Partisanscan self-select like-minded content at either end of the political spectrum, and thusstrong partisans on both sides are exposed primarily to like-minded politicalperspectives in their newspapers.

In the 1990s, there was limited partisan choice among mainstream dailynewspapers in the United States. At least at that time, as shown in figure 5-1,Republicans found their daily newspapers significantly less compatible with theirpolitical views than Democrats did. Relative to newspaper readers in Great Britain,

228 diana c. mutz

Figure 5-1. Exposure to Like-minded Media Content by Partisanship andMedium, 1996a

1

2

3

4

5

6

Strong Republican Independent Strong Democrat

Newspapers

Television news

Talk shows

Extent of exposure to like-minded content

Source: Based on the data shown in Mutz and Martin (2001, figure 4A).a. Respondents were asked about their most frequently used source in each category.


Americans were not as successful at selective exposure. Most Republicans found lessto like about the partisanship in television news broadcasts than did Democrats.Not surprisingly, Republicans found talk shows to be highly compatible,whereas Democrats found little to comport with their views on talk shows.10

Perhaps Americans exercised selective exposure by gravitating to differentmedia (Republicans to talk shows, Democrats to television) rather than findingsomething politically compatible within each medium. But one wonders howdifferent these patterns would look if the same data were available for 2006.Since the 1990s, Fox News has provided Republicans with a politically compat-ible television network, and the number of news sources on television and theInternet has proliferated. In addition, the availability of openly partisan tele-vision programs, radio shows, and websites may mean that the pattern of U.S.news consumption will more closely approximate that of British newspapers inthe not-so-distant future. To the extent that the media environment in theUnited States continues to develop along the lines of a niche market based onpartisanship, this new news environment should facilitate selective exposure bymaking it easier for people to find the most compatible source of news.

Be that as it may, data demonstrating that selective exposure to media isresponsible for mass polarization do not yet exist. Nonetheless, a recent experi-mental study of Internet news attempted to mimic the kinds of choices peoplemake in today’s media environment. Shanto Iyengar and Richard Morin attemptedto document evidence of partisan-based selective exposure by randomly assigninga set of news stories to four news sources: Fox News, National Public Radio(NPR), CNN, and the BBC.11 Participants were shown four news story headlines,one under each news organization’s logo. Participants were then asked which ofthe reports they would like to read. This scenario was repeated for articles in sixnews categories, including American politics, the war in Iraq, race in America,crime, travel, and sports. Some respondents were randomly assigned to a controlcondition without logos of any kind attached to the stories.

The results suggest that selective exposure is alive and well—even among rel-atively mainstream news sources. Republicans overwhelmingly preferred storiesfrom Fox News, whereas Democrats divided between NPR and CNN. Democratssystematically avoided Fox News, whereas independents demonstrated no par-ticular pattern of preferences. And while the difference between Republicans


10. Mutz and Martin (2001, pp. 105–07).11. Shanto Iyengar and Richard Morin, “Red Media, Blue Media: Evidence for a Political Litmus

Test in Online News Readership,” Washington Post, May 3, 2006.


and Democrats was most pronounced when the stories involved hard news(such as national politics and the war in Iraq), it did not disappear entirely evenwhen the story was about possible vacation destinations.

A comparison of how many partisans chose certain stories with news organi-zation labels versus the same stories without labels provides a clear picture ofhow news sources may serve as a guide to selective exposure. Republicans werethree times as likely to choose a news story accompanied by the Fox News logoas without it, and less likely to choose it if it was labeled a CNN or NPR story.Among Democrats the pattern was less pronounced; they were somewhat morelikely to choose a story if it was labeled from CNN or NPR and significantly lesslikely to choose the same story labeled as being from Fox News. As news con-sumers increasingly have the ability to customize their news environments totheir own tastes, the likelihood that news will simply reinforce existing viewsand produce a subsequent polarization of partisan groups seems all the moreplausible.

Relative to much of the twentieth century, the current direction of politicalnews seems to portend greater division rather than greater consensus. But in themidst of considerable alarmism about the impact of modern media, it is worthremembering that during the days of the openly partisan press of the nineteenthcentury, selective exposure was even easier to accomplish. Newspapers of thatera openly aligned themselves with political parties and endorsed candidates onevery page, not just in their editorial content.

But even though the question of whether selective exposure exists seems tohave been resolved affirmatively, the consequences of selective exposure for publicattitudes have not yet been well studied. It is common to assume that if Democrats,Republicans, and independents all exposed themselves to exactly the same infor-mation, the public would be less polarized. Perhaps, but some evidence suggeststhat this would not happen. Macro-level research on selectivity in the processingof political information suggests that even if audiences did experience identicalexposure to political information, it might not diminish polarization muchbecause of the selective way in which partisans interpret new information.

Imagine a situation in which new information becomes available to all threegroups—for example, news that the economy has improved. Assume also thatthere is a Republican president in office, so that the president’s job approvalstarts out higher among Republicans than among independents, who in turn givethe president more favorable job approval ratings than Democrats do. Nonetheless,all three partisan groups in a situation such as this generally increase their levelsof presidential approval. In other words, they respond to new information in a

230 diana c. mutz


reasonable manner. Likewise, downturns due to bad news such as declining eco-nomic indicators would generally cause all three groups’ approval levels to fall.According to aggregate data compiled over time, Republicans, Democrats, andindependents do indeed respond by changing their views in the same direction—and to roughly the same extent—in response to new information.12 But this kindof parallel change is not particularly helpful in reducing polarization. If eachgroup moves up or down to roughly the same extent, the gap between Republicansand Democrats remains the same. The two groups are as polarized as beforethey received new information.

The underlying assumption of much work on information processing is thatif people were truly unbiased processors of the same political information, theyshould ultimately converge in their judgments. As new information becomesavailable (perhaps news that the economy has worsened), partisans update theirpresidential approval ratings in light of their initial views. The extent of decreasedpresidential approval due to negative information should not be even across allgroups, but should be more pronounced in groups that begin with higher levelsof approval. New information matters more when it contradicts initial expecta-tions. In this scenario, new information should logically result, over time, in aconvergence of opinion. Whether the news is positive or negative, the threegroups’ opinions should move closer and closer together.

Empirical data suggest, however, that convergence is not usually what hap-pens. Instead, patterns of aggregate opinion suggest that partisanship is a drivingforce in how people perceive, interpret, and respond to the political world.13

This means that even if selective exposure to information were not a problem, it isdoubtful that information from mass media would decrease the gap between theperspectives of Republicans and Democrats. Biased processing is prevalent inthe American public, and partisanship is its driving force.

Biased processing models underscore a skepticism prevalent in contemporarypolitical psychology that information is the cure-all for what ails the quality ofpolitical decisions. If people are not passive recipients of information, but ratheractive choosers, interpreters, and rationalizers, then mass media are severely lim-ited in what they can do about political polarization. Even if citizens were stilllimited to the choice between three indistinguishable network news broadcasts,biased assimilation models would predict that differences would persist becauseof the ways in which people process the information they receive.


12. Gerber and Green (1999).13. Bartels (2002).


The Internet has become a particular target of concern with respect to selec-tivity. Law professor Cass Sunstein, author of Republic.com, has suggested that thecacophony of online political voices, including some very extreme ones, raisesthe possibility that extreme political views will be reinforced and encouraged.14

Selectivity is made easier on the Internet by search engines and links betweenwebsites that espouse similar views. The voices represented on the Internetinclude those of fringe groups whose extremist ideas would never be covered bymore traditional mainstream media. For many, that alone is cause for concern.While it is possible (even likely) that these groups have always been out there, theyare more visible and better organized now as a result of the Internet. And even iftheir numbers have not increased as a result of the Internet, better organizationmeans that they could be more likely to pose a threat by taking violent action ofsome kind. Still others worry that their mere presence on the Internet encouragesthose with similar political leanings by reinforcing and exacerbating the extremityof their views, and by convincing them that they are not alone and thus neednot abandon their unpopular positions. In all of these scenarios, Internet-basedselective exposure is cause for alarm.

Nonetheless, the irony of all of this hand-wringing about the Internet andthe expansion in the number of political voices on television should not gowithout notice. For years, scholars who studied mass media voiced concernsabout the near monopolistic media and the paucity of non-mainstream politicalvoices on the air or in print. Today, the marketplace is more—and less—ideal.The proliferation of voices on the Internet means there is a larger marketplaceof political ideas than in the past, but that does not necessarily lead citizens tofreely sample all of its products.

Choosing Whether to Watch Political News

Partisan-driven selectivity is only one mechanism by which increasing mediachoice has been tied to increasing polarization. In a more novel argument link-ing choice and polarization, Markus Prior suggests that greater media choicein the form of entertainment content, combined with the availability oftwenty-four-hour news channels, has widened the gap between the politicallyinformed and those without interest and information.15 The rise of cable televi-sion has clearly changed contemporary presidents’ ability to commandeer thenational airwaves, as political scientists Matthew Baum and Samuel Kernell

232 diana c. mutz

14. Sunstein (2001).15. Prior (forthcoming 2007).


have shown.16 The average size of the audience watching prime-time presidentialaddresses and news conferences decreased steadily in the late twentieth century—a trend that appears to be primarily a function of the availability of cable televi-sion. Likewise, debate watching fell, particularly among cable viewers. Peoplewithout cable are essentially a captive audience, and their debate watching is lessa function of political interest than it is for those with cable, for whom watchingthe president on television is clearly a choice.

Many of today’s television viewers no longer need to suffer through a State of theUnion address if they would rather watch a sitcom. In the era of hundreds of chan-nels from which to choose, viewers need not watch the political debates or presi-dential press conferences that dominated all three networks not so long ago. Thereis always another—and often far more entertaining—choice for the less politicallyinterested American, from cooking shows to crime dramas to feature films.

The lack of largely involuntary exposure to political content among the lesspolitically involved, according to Prior, makes this group even less likely to turnout to vote than before. (On the other hand, there is a steady diet of politicalmedia for political junkies, thus further energizing and reinforcing their proclivitiesto be politically active.) Thus, changes in the media environment lead to less invol-untary exposure to political information and less incidental learning about thepolitical world. This “byproduct learning,” it is argued, is what motivates the lesspolitically interested to turn out to vote. Without it, the increasing gap in politicalknowledge between the politically interested and uninterested translates into anincreasing turnout gap at the polls. Less politically involved, more moderate votersare less likely to turn out than before, while the more politically involved are evenmore likely to vote than usual. The electorate is thus robbed of its middle, with theresult that the electorate that does go to the polls is more polarized in its views.17

In light of this long chain of proposed causal influences, Prior’s evidence insupport of this theory consists of tests of the many links that make up this largerprocess. Using an experiment embedded in a national survey, he randomlyassigned respondents different amounts of choice, asking the hypothetical ques-tion, “If you had free time at 6 o’clock at night and the following programs wereavailable, which one would you watch, or would you not watch television then?”Respondents in the low-choice condition were given only five options: the three


16. Baum and Kernell (1999).17. Notably, this process results in no individual-level shifts in extremity of views, and thus in

no polarization at the level of individuals or the mass public. But it produces a compositional shiftin the electorate such that the extremes are well represented in voting booths, but the middle, lessextreme segments of the population are less likely to be heard.


network news programs, The NewsHour with Jim Lehrer, and not watching atall. Respondents in the high-choice condition had those five options, plus severalcable news programs and several entertainment options described as “a comedy orsitcom program like Friends or The Simpsons,” “a drama program like ER or Lawand Order,” and similar examples from other genres including science fiction,reality TV, and sports.

The results of Prior’s tests suggest that in the high-choice condition the newsaudience is half of what it is in the low-choice condition. To the extent that thishypothetical situation generalizes to real-world changes in U.S. media, this is astriking result. But it also raises the question of why so many people watched thenews in the past when it was not what they ideally would have selected. Mediaeconomists suggest that they did so out of habit, as well as a tendency to decide towatch television before knowing precisely what the viewing options were. Thisprocess no longer rings true for many Americans today. As the ability to “timeshift” the viewing of preferred programs becomes more popular, and the degreeof choice at any one time becomes still more expansive, it seems unlikely thathabitual viewing times will result in much incidental exposure to political content.

A related—and important—matter is whether the politically unenthusedreally picked up political knowledge from their incidental viewing in the days ofless choice. And what about the pre-television era, when the literacy levels of theless educated segments of the population limited their direct access to politicalinformation? Prior suggests that during the pre-television era, when media optionswere largely limited to newspapers and radio, a similar gap existed between thepolitical haves and have-nots, but that television exerted a democratizing influenceon the public, making it less effortful for even the less educated to receive politicalnews. The less interested still remained less knowledgeable than in the days beforetelevision, but the gap was smaller.

Using National Election Studies data from the 1950s and 1960s, Prior docu-ments how the number of television stations to which an individual had accesspredicts political knowledge, particularly among the less educated segments ofthe population. His models suggest that access to television has increased politicalinterest and turnout among the less educated segments of the population.

With the more recent advent of cable television and the Internet, the gapbetween the politically informed and uninformed should be widening again in thecontemporary population. To measure this, Prior develops the concept of “relativeentertainment preference”—that is, the extent to which a given respondent prefersentertainment over news. One measure involves directly asking people whatgenre they like best from a list of ten categories, then having them select a second,

234 diana c. mutz


third, and fourth favorite. Of the remaining options, they are allowed to indicatethose that they “really dislike.” The respondents who claim to actively dislike newsare at the high end of the “relative entertainment preference” scale, while those whoindicate neither a like nor a dislike are next highest, and those who rank newsamong their four favorite types of programs are at the lower end of the scale.

Central to Prior’s thesis is the group he dubs “switchers,” people who willswitch from news viewing to entertainment viewing when given the choice. Thisis the segment of the population whose level of political knowledge (and level ofturnout at the polls) should be most affected by greater choice. Because switchersabandon news once they have a choice of entertainment content, they reducetheir incidental exposure to political content, which in turn limits their accidentalpolitical learning.

Prior’s thesis presents a conundrum for a society that is addicted to choice.Some segments of the population will expose themselves to political news vol-untarily and turn out to vote regularly regardless of the shape of the current newsenvironment. Others will not watch political news regardless of how little choicethey have, and they are unlikely to bother voting in any case. But for those on themargins of these groups—those who are not political activists, yet not totallydisengaged—incidental exposure to the political world is an important way inwhich they are drawn into the political process. Without involuntary exposureto political media, these more moderate voters will drop out of the electoralprocess, and the voting public will be increasingly extreme in its composition.

Losers’ Consent in an Age of Illegitimacy

Yet another way in which media may contribute to political polarization stemsfrom the nature of contemporary media content during elections. It has beenoften observed, extensively documented, and widely lamented that election cov-erage tends to emphasize the “behind-the-scenes” insider perspective on whatdrives election outcomes. Rather than emphasize the substantive reasons whyvoters might have selected one candidate over another, media coverage tends toemphasize the role of political consultants, advertising campaigns, and otherfactors that have less to do with the quality of the candidate or the extent towhich aspects of his agenda resonate with voters, and more to do with strategicaspects of how campaigns are waged.

In his book Out of Order, Thomas E. Patterson documented the stark increasein this type of coverage from the 1960s through the 1990s.18 Patterson’s analysis


18. Patterson (1993).


shows that it is not simply that horse-race coverage—that is, coverage of who isahead and who is falling behind—remains popular among journalists. (Indeed,many argue that horse-race coverage is a time-honored journalistic tradition thatpredates both polls and television.)19 But in comparison with forty or fifty yearsago, there is now far more of what Patterson calls “game-centered” coverage ofelections, in which journalists focus on the suspense and speculation about likelywinners and losers and the behind-the-scenes machinations of candidates andtheir staffs. As Patterson notes:

The game schema dominates the journalist’s outlook in part because itconforms to the conventions of the news process. . . . The conventions ofnews reporting include an emphasis on the more dramatic and controversialaspects of politics. . . . The game is always moving: candidates are continuallyadjusting to the dynamics of the race and their position in it. Since it canalmost always be assumed that the candidates are driven by a desire to win,their actions can hence be interpreted as an effort to acquire votes. Thegame is thus a perpetually reliable source of new material.20

Patterson’s analysis of New York Times coverage between 1960 and 1992demonstrates that in 1960 the majority of election stories were framed as storiesabout candidates’ policy positions. By 1992, election stories were six times morelikely to be framed in the context of campaign strategy than as policy stories.Over the course of several decades, stories about candidates’ policy positions werereplaced by an increasing number of articles about the campaigns themselves.

In what comes across as an effort to give ordinary people a behind-the-scenesperspective on how elections are run, today’s election coverage is replete withanalyses of the strategic decisions involved in campaigning for office. When apresidential candidate gives a speech about health care in Atlanta, campaignjournalists are likely to frame the event not in terms of a discussion of health carepolicy or the candidate’s platform, but as an effort to woo southern voters. Forreporters—and perhaps for their audiences as well—the text is not nearly asinteresting as the subtext.

Some speculate that journalists highlight the strategic aspects of campaigns asa means of insulating their readers and viewers from the more manipulativeaspects of campaigns; if audiences are aware of candidates’ real intentions andthe reasons for their actions, then surely they will not be duped by them. In hisbook Seducing America, Roderick P. Hart has suggested that strategic coverage is

236 diana c. mutz

19. Sigelman and Bullock (1991).20. Patterson (1993, pp. 60–61).


popular with the public because it gives them the illusion of being informed andknowing what is really happening in the campaign.21

The problem with constant reporting on the campaign is that it suggests tonews media audiences that being persuaded by a politician is a bad thing, a sign ofweakness, gullibility, or stupidity. For the same reason that no one admits to havinghis or her mind changed by a political ad, few Americans see open-mindednessto campaign propaganda as a good thing. Another problem with game-centriccoverage is that it shifts the audience’s attention away from what is being said(on health care, for example) toward the journalist’s perspective on the strategicreason why it is being said.

From the perspective of polarization, why is the rise in this kind of coverage aconcern? The link to polarization comes later in the election process, after the win-ners and losers have already been decided. To the extent that citizens are con-vinced by game-centric coverage that elections are won or lost by the right choiceof advertisements, a new haircut, the best speechwriters, or the cleverest consultants,one could understand why voters on the losing side in any given contest mightfeel they have been robbed. The best advertisements do not necessarily mean thebest presidents, and a good haircut seldom leads to good governance. Moreover,a good campaign can be waged by a weak candidate, and a bad campaign by astrong candidate.

It is a much more polarizing experience to think that one’s candidate lost anelection because of a minor misstatement, a failure to look good on television,or a failure to hire the right advertising agency than to think that he lost becausehis policy ideas were less popular than his opponent’s. If electoral victories are,according to the press, about the candidate’s ability to play the game, then electorallosses also are about the candidate’s failure to play the game well.

Empirical evidence backs up this broad argument, though scholars have yet totest the multiple steps in this process within a single study. For example, there isconsiderable empirical evidence of the media’s increased emphasis on the strategicaspects of campaigns. A separate body of evidence suggests that the explanationsthe mass media put forth for how an election was won or lost are widely adoptedby the public. Today’s media coverage is more likely to suggest apolitical, illegit-imate explanations for a candidate’s electoral victory, and the public on the losingside is likely to believe them.22 A belief that one’s side has experienced an illegit-imate loss, in turn, prompts the losing partisans to become increasingly angry


21. Hart (1999, pp. 77–100).22. See Mutz (1993); Hershey (1992).


and frustrated. If one believes the other side won by running deceitful ads, thenit is easy to villainize the opposition and thus create more extreme perceptionsof the consequences of one political choice over another. At this point, it is nolonger about differing political philosophies; it is about right versus wrong,truth versus deceit, good versus evil.

The link in this chain of events that has not been well documented is the extentto which delegitimizing explanations lead to more polarized views, particularlyin the form of an intensified dislike for the “other side.” In a 2005 book, Losers’Consent, Christopher J. Anderson and his coauthors examine the role that electionsplay in legitimizing their own outcomes. For citizens who end up on the winningside in a given contest, accepting the outcome is not particularly difficult. Butwhat produces consent from the losers? Not surprisingly, those who participatein an election but end up on the losing side are significantly more negativeabout their system of government. Losing naturally leads to anger, particularly ifthe loss is seen as illegitimate—and evaluations of the legitimacy of elections anddemocracy are significantly less positive among losers than among winners.23

Although it has yet to be tested, one would likewise expect that thoseexposed to media coverage that emphasizes delegitimizing explanations of anelection outcome—such as the idea that the election was won primarily by adscontaining lies or half-truths—should come away from the experience of lossmore polarized in their views. Nobody wants to end up on the losing side, butlosing to an opponent who won for the wrong reasons would be expected to cre-ate much more negative feeling toward the opposition.

By suggesting this, I do not mean to refer specifically to the unusual, extremelyclose election outcomes that occurred in the 2000 and 2004 presidential elections,though they obviously come to mind. The more general point is that regardless ofhow lopsided an outcome might be, media coverage plays an important role instructuring explanations for why one side was victorious and the other side failed.Once the people have spoken, it is the mass media that decide what they havesaid. To the extent that delegitimizing explanations are encouraged by the pressand adopted by the losing side, a more polarized electorate is likely to be the result.

Media Bracketing of Acceptable Opinion

Yet another explanation linking the media to political polarization originatesin two studies of the role of the news media as a political institution. In a 1984

238 diana c. mutz

23. Anderson and others (2005).


study, political scientist Dan Hallin examined the oft-repeated assertion thatduring the war in Vietnam the American news media (television in particular)became more oppositional toward government and thus helped lead publicopinion away from support for the war. In Hallin’s study of political elites,media coverage, and mass opinion, he found little support for this thesis of anoppositional media. Although media coverage of the war did become increasinglycritical after the Tet Offensive in 1967, Hallin demonstrated that this was becauseelite opinion became more divided on the war at that time. By relying on elitesources, reporters mirrored the division of elite opinion—and subsequentlytheir coverage influenced the range of opinions held by the mainstream publicas well.24

Hallin’s study pointed to the importance of the mass media in “bracketing”the range of acceptable opinion for the public. By defining the boundaries ofacceptable controversy, the media define the range of legitimate opinions thatthe public may adopt. An issue such as Vietnam may shift over time from thesphere of consensus to the sphere of legitimate controversy, but this is a top-down phenomenon originating with elites rather than with the media. Othershave similarly noted that media coverage tends to be “indexed” to the range ofelite opinion rather than more broadly to the range of mass opinion.25

When members of Congress represent some of the more extreme positions(as they do now), it follows then, as a result of journalists’ tendency to rely onofficial sources, that those more extreme viewpoints are also more likely to becovered by the press. Although the views that are covered originate with elites(such as members of Congress), the media play an important role in conferringlegitimacy on a given range of views simply by covering those perspectives regu-larly, but not others. What this process suggests is that by covering what isessentially a more polarized group of political elites, journalists may be offeringa wider range of acceptable views to the public and thus discouraging consensus.If the smorgasbord of views offered in today’s media represents greater extremesthan in the past, then it is perhaps not surprising that more of the public nowendorses more extreme views.

This “indexing” or “bracketing” hypothesis has been relatively well documentedin terms of the effects that political elites have on the range of views reported inmainstream media. But the impact of indexing on the mass public within thosemedia audiences is less well studied. By simply broadening the range of views


24. Hallin (1984).25. The use of “indexing” in this context was coined by Bennett (1990).


they present, does media coverage also produce a public with more diverseviews? The jury is still out with respect to the impact on the public.

To the extent that this mechanism of polarization holds true, it suggests thatperhaps the media critics of the previous generation should have been morecareful about what they wished for. In a paradoxical way, longtime critics of thelack of diversity in media and the lack of representation for more extreme voiceshave seen the change they desired: there is no question that today’s media carrya broader range of political perspectives than they did a generation ago. But thevery diversity of those same voices may encourage the public to adopt morepolar positions—an unforeseen and less desirable consequence of a more vibrantmarketplace of political ideas.

In-Your-Face Politics: Gut Reactions to Televised Incivility

“Shout-show” television has been the target of a tremendous amount of criticismfrom many quarters, academic and otherwise. The world of political disagreementas witnessed through the lens of political talk shows is quite polarized. Increasedcompetition for audiences has led many programs on political topics to liven them-selves up in order to increase audience size.26 Thus, political talk shows such asThe McLaughlin Group, The O’Reilly Factor, Meet the Press, Capital Gang, Hardball,and many others tend to involve particularly intense and heated exchanges.

The issue of contentiousness and incivility in political discourse was brought toa head in October 2004, when Daily Show host Jon Stewart appeared on Cross-fire and openly criticized this program (and others like it) for its “partisan hack-ery,” which Stewart said was “hurting America.” Is there any truth to Stewart’sclaim? Aside from the obvious distastefulness some find in watching politiciansscream, yell, and interrupt one another for thirty minutes or more, how is thiskind of in-your-face politics implicated as a potential cause of mass polarization?

The tendency on television is to highlight more emotionally extreme and lesspolite expressions of opinion, and research suggests that these expressions ofincivility may have important consequences for attitudes toward the opposition.These consequences flow from the fact that politeness and civility are more thanmere social norms; they are means of demonstrating mutual respect.27 In otherwords, uncivil discourse increases polarization by helping partisans think evenless of their opponents than they already did.

240 diana c. mutz

26. Fallows (1996).27. This is a point made in Kingwell (1995).


And yet market forces seem to favor the kind of television that encouragespolarization. Polarized political discourse and an angry opposition makes forcompelling television. Viewers may claim that they find it disgusting, but theycannot help watching—just as passing motorists cannot help “rubbernecking”when there is an accident alongside the highway. It is not that people actuallyenjoy what they are seeing, but there is something about information of thiskind—information about life and death, about conflict and warring tribes in adispute—that makes it difficult to ignore. Evolutionary psychologists havepointed to the adaptive advantage of having brains that automatically pay attentionto conflict as a means of staying alive in an earlier era. At a cognitive level, ofcourse, no one really expects to be caught in the “crossfire” of a televised partisanshout-fest. But even when it is “only television,” and thus poses no real threat ofbodily harm, people cannot help but watch and react to incivility.

My own research suggests that psychologists are correct about the demandsof incivility on human attentional processes. To examine the difference thatincivility makes independent of political content, I produced a mock politicaltalk show—on a professional television set using professional actors as congres-sional candidates. The candidates espoused the same issue positions and madeexactly the same arguments for and against various issue positions in two differentversions of the program. In one discussion, however, they raised their voices, rolledtheir eyes, and engaged in an impolite, uncivil exchange. In the civil version ofthe program, they spoke calmly, refrained from interrupting one another, andshowed mutual respect simply by obeying the social norms for polite discourse.28

The differences in viewer reactions to the two programs were startling. Thegroup randomly assigned to the uncivil version of the political discussion cameaway with roughly the same feelings toward their preferred candidate as those inthe civil group. But attitudes toward the “other side” became much more intenselynegative when the two exchanged views in an uncivil manner. The more dramatic,uncivil exchanges encouraged a more black-and-white view of the world: theircandidate was not just the best; the alternative was downright evil.

This effect was evident for partisans on both sides of the political spectrumand regardless of which candidate they liked best. Interestingly, watching theuncivil version led to greater polarization in perceptions of “us” versus “them,”relative to a control group, but watching the civil version of the exchange led todecreased levels of polarization. This pattern of findings suggests that politicaltelevision has the potential to improve as well as to exacerbate the divide among


28. Mutz and Reeves (2005).


partisans of opposing views; it simply depends upon how those differences ofopinion are aired. When differences of opinion are conveyed in a manner thatsuggests mutual respect, viewers are able to understand and process the ratio-nales on the other side and are less likely to see the opposition in starkly nega-tive terms. Differences of opinion are perceived as having some legitimate andreasonable basis. But when those same views and rationales are expressed in anuncivil manner, people respond with an emotional, gut-level reaction, rejectingthe opposition as unfairly and viciously attacking one’s cherished views.

Using indicators of physiological response, my studies also demonstrate thattelevised incivility causes viewers’ levels of emotional arousal to increase, just asthey do when people encounter face-to-face incivility. In the face of real-worldconflict, this reaction supposedly serves a functional purpose—participants aregiven the rush of adrenaline they may need to flee the situation. But with tele-vised incivility, this kind of reaction serves no purpose; it is simply a remnant ofbrains that have not adapted to twentieth-century representational technology.29

Even though viewers are just third-party observers of other people’s conflictson television, they show heightened levels of emotional arousal, just as peopledo when encountering face-to-face disagreement. This is not so surprising if oneconsiders how it feels to be a third-party observer of a couple’s argument at adinner party. The same discomfort, awkwardness, and tension exist, even forthose not directly involved in the conflict. Likewise, when political commentatorRobert Novak stormed off the set of a live broadcast of CNN’s Inside Politics inAugust 2004, viewers were uncomfortable—and they paid attention. The tensionwas palpable to viewers, even though few may be able to remember what thesubstance of the conflict was.

The heightened arousal produced by incivility can make it difficult to processthe substance of the exchange. Some arousal helps to call attention to what other-wise might be considered bland and uninteresting. But at extremely high levelsof arousal, people will remember only the emotional content of the program(who screamed at whom, who stomped off in a pique) and recall little of thesubstance of the disagreement. As anyone who has ever had an argument knows,there is a point at which the emotional content of the exchange overwhelms anypotential for rational discourse. As a result, viewers gain little understanding ofthe other side. They perceive their own side of the debate as unfairly attacked,and thus the incivility their own candidate displays is simply an appropriatelevel of righteous indignation in reaction to an unprovoked attack. The incivility

242 diana c. mutz

29. Reeves and Nass (1996).


demonstrated by the opponent demonstrates that he is a raving lunatic, whollyunfit for office.

In addition to this disdain for the opposing side, incivility produces a secondimportant reaction—heightened attention. As Bill O’Reilly, host of The O’ReillyFactor, suggests, “If a radio producer can find someone who eggs on conservativelisteners to spout off and prods liberals into shouting back, he’s got a hit show.The best host is the guy or gal who can get the most listeners extremely annoyed—over and over and over again.”30 Evidently, these sorts of shows have hookedSenator Hillary Rodham Clinton (D-N.Y.), who indicated that she and herhusband Bill now have TiVo, a technology that allows a viewer to record andreplay television programs. And for what purpose do the senator and the formerpresident use TiVo? According to Senator Clinton, they use it to record the mostoutrageous statements made by their political opponents so they can play themover and over and yell back at the television.31 An optimist might regard thisvignette as an example of how viewers are not necessarily selectively exposingthemselves to politically compatible media. But the pessimist would undoubtedlypoint out that yet another media mechanism of polarization has kicked in totake its place. Uncivil political discourse that produces such strong emotionalreactions is unlikely to further the cause of political moderation.

Controlled laboratory studies suggest, for better or worse, that O’Reilly iscorrect: incivility is extremely entertaining and people like to watch it, even if itis just to scream back. Despite the fact that many viewers claim to be repulsedby it, the respondents who viewed the identical but uncivil version of the sameprogram always rated it as more entertaining, found it more exciting to watch, andindicated a greater desire to see the uncivil program again than the civil version.32

Polite conversation is boring, and the deliberative ideal for political discoursemakes for dull television. “I acknowledge there are some good points on myopponent’s side” will probably never make good television, whereas “These evilpeople must be stopped!” always will.

With these findings in mind, it is important to consider the extent to whichthe rise of televised political incivility can help explain mass polarization. Ispolitical discourse truly any more uncivil now than in the past? Some have sug-gested that the United States is in the midst of a “civility crisis” in its public life.


30. O’Reilly (2000, p. 52).31. Senator Hillary Rodham Clinton, speech at the Doherty-Granoff Forum on Women Leaders,

Brown University, April 8, 2006.32. Mutz and Reeves (2005).


As then University of Pennsylvania president Judith Rodin argued in 1996,“Across America and increasingly around the world, from campuses to the halls ofCongress, to talk radio and network TV, social and political life seem dominatedtoday by incivility. . . . No one seems to question the premise that politicaldebate has become too extreme, too confrontational, too coarse.”33 Similar callsfor greater civility in political discourse have come from a wide array of scholars,as well as from philanthropic organizations.

Some scholars concur that incivility is on the rise. Political scientist Eric M.Uslaner, for example, has suggested that members of Congress are increasinglylikely to violate norms of politeness in their discourse.34 Linguist Deborah Tannenhas characterized the United States as having “a culture of argument” thatencourages “a pervasive warlike atmosphere” for resolving differences of opinion.35

Even journalists concur that “hyperbole and venomous invective have becomethe order of the day” in American politics.36

Clearly, there is a widespread perception that political discourse is much moreuncivil now than in the past, but there is little historical evidence to confirmsuch a trend. As then senator Zell Miller (D-Ga.) implied when he wistfully saidhe would like to challenge Hardball host Chris Matthews to a duel, violenceamong political opponents was once far more common than it is now. SenatorMiller’s statement was made during an uncivil exchange between himself and ajournalist during the 2004 Republican National Convention. It made headlines,precisely because the idea of using weapons to resolve political differencesseemed absurd. We have not had a duel to the death among politicians for manyyears, and thus one could easily characterize today’s political talk shows as mildby comparison.

So is it fair to say that incivility is on the rise in political discourse? There is nodefinitive answer to this question, but the increased visibility of uncivil conflictson television seems indisputable. Although politicians of past eras may frequentlyhave exchanged harsh words, without television cameras there to record theseevents and to replay them for a mass audience their impact on public perceptionswas probably substantially lower. The dominance of television as a source ofexposure to politics suggests that public exposure to uncivil political discourse has

244 diana c. mutz

33. Judith Rodin, speech at the Penn National Commission on Society, Culture, and Community,December 9, 1996 (www.upenn.edu/pnc/pubkeynote.html).

34. Uslaner (1993).35. Tannen (1998).36. Michiko Kakutani, “Polarization of National Dialogue Mirrors Extremists of Left and

Right,” New York Times, November 26, 2000.


increased. Moreover, it is one thing to read about political pundits’ or candidates’contrary views in the press, and quite another to witness them directly engagedin vituperative argument. The sensory realism of television conveys a sense ofintimacy with political actors that people were unlikely to encounter in the past,even among the few lucky enough to have face-to-face meetings.

Television provides a uniquely intimate perspective on conflict. In the literatureon human proxemics, the distance deemed appropriate for face-to-face inter-actions with public figures in American culture is more than twelve feet.37 Yetexposure to politicians on television gives the appearance of being much closer.When people are arguing, the tendency is to back off and put greater spacebetween those who disagree. Instead, when political conflicts flare up on television,cameras tend to go in for tighter and tighter close-ups. This creates an intenseexperience for the viewers, one in which they view conflict from an unusuallyintimate perspective. Political scientist Jane J. Mansbridge has noted that whenopen political conflict occurs in real life, bringing people together in one another’spresence can intensify their anger and aggression.38 To the extent that a televisionpresence has similar effects, incivility is likely to encourage polarization.

What Are the Prospects?

In thinking about the prospects for the future, it is useful to consider separatelyand collectively the four mechanisms by which the media may be involved in thepolarization of the mass public. Two of these mechanisms, discouraging losers’consent and bracketing the range of acceptable opinion, are driven by the contentof news media. In one case, journalists’ game-centric coverage of campaigns isimplicated, and in the other, the diversity of political voices affects the range ofmass opinion. The other two mechanisms are driven more or less by technologicalchange. Because of the proliferation of channels on cable television, along withthe development of the Internet, most Americans have access to more partisanpolitical voices than they ever did before. Whether through partisan-based selectiveexposure, or through selective avoidance of political content altogether, the dis-tribution of opinion in the electorate has changed. Finally, the ability of televisionto present an unnatural “in-your-face” perspective on politicians may have furtherincreased levels of vitriol directed toward the opposition.


37. Aiello (1987).38. Mansbridge (1983).


How likely is it that these processes will change? Theoretically, the mechanismsbased on media content could become less tenable if journalistic norms were tochange. For example, if not only Crossfire were taken off the air, but also everyother uncivil shout-fest show that involves politics, then one could see how thispolarizing mechanism might abate in significance. Likewise, if campaign cover-age were to shift away from its predominantly strategic angle and focus insteadon taking political statements at face value, one could imagine losers who wouldbe less angry about the opposition’s victory, particularly if it were deemed toresult from legitimate differences of opinion. There are several problems withsuch proposed changes, of course. Even if they were possible to implement(which they are not), it seems unlikely that they would offer tenable solutions oreven modest improvements on the current situation.

News is commerce, to be sure. What all four of the mechanisms described inthis chapter share is a tension between drawing viewers, readers, and listeners topolitical media, and producing media content that does not encourage polarizedpolitical perspectives. The goal of niche marketing and the extent of choice is togive viewers more of what they want, but not necessarily more of what theyneed. Likewise, the game-centric, campaign-oriented journalism that is so commontoday exists because viewers and readers find such discussions of interest. Thegreater publicity that extreme views now receive probably also helps draw largeraudiences. It is far more dramatic to show starkly opposed foes engaged in battlethan to illustrate the political world with two moderates calmly discussing theirsmall differences. If it were not for the need to make politics more interestingfor marginally politically involved viewers, then it is doubtful that there wouldbe a need for confrontational shout shows to exist either.

Moreover, any changes to media content designed to alleviate the impact ofone of these mechanisms are likely to replace one mechanism of media polariza-tion with another. Take, for example, the incivility that is so prevalent in politicaltalk shows. What would happen if producers were to tone down these programsso that levels of civility approached the norms of everyday people in their socialinteractions? On the one hand, viewers would not have the chance to intensifytheir dislike for the opposition by viewing uncivil repartee. But on the otherhand, in-your-face incivility is precisely what people like about these shows. Instudy after study, even those who responded negatively to incivility reportedthat the uncivil shows were more interesting and exciting to watch. Thoseexposed to the civil versions of the same show were far less likely to indicate anyinterest in seeing it again or in watching it regularly. As a result, in a hypotheticalera of civil political television, those viewers would be unlikely to watch at all.

246 diana c. mutz


They would fall into the category of viewers that Prior refers to as “switchers”—people who would switch to entertainment when given a choice—and thus fallout of the reach of political information altogether.

Three general approaches to improving this situation are possible. The first is topush for less commercial pressure. In Breaking the News, James Fallows excoriateshis fellow journalists for turning news coverage into more of an effort to entertainthan to edify. He faults the focus of media companies on their bottom lines as thesource of the corruption of news.39 From a broader perspective, however, it seemstoo facile to claim that the commercial interests of U.S. media companies are toblame. What Fallows misses is that even if all political content were government-funded, the lack of commercial pressures would not alleviate the need to attractaudiences to political content. In order to do good or harm, political news musthave large audiences; high-minded programs with small audiences are hardlybetter than no programs at all.

A second, equally fruitless, approach has been to wag fingers accusingly at thecivically bankrupt American public. Why is it, many ask, that Americans do nottake their civic duty more seriously and educate themselves on political issues?

A third option is to change the shape of political media so that they can drawaudiences without inadvertently polarizing the public. Traditional televisionnews programs are in a state of turmoil these days over which direction to take.Do they reinvent themselves to become more like entertainment programs and viefor audiences on the edge of political interest? Or do they take the so-called highroad and remain serious and hard-hitting? The problem with the high road isthat even if it leads to critical acclaim and journalistic respect, it is unlikely to drawin the politically marginal audiences who prefer to watch entertainment. Bymany normative accounts, doing the “right thing” would result in precisely thewrong effect—further chipping away at the proportion of Americans who exposethemselves to politically relevant television content.

In a 2005 interview, CBS chairman Leslie Moonves indicated just how drasti-cally he thinks traditional news broadcasts must be revamped in order to increaseaudiences and survive. “We have to break the mold,” he stated. The alternativesMoonves has considered include mold-breaking formats ranging from the com-edy of Jon Stewart’s mock news program, The Daily Show, to the zany humor ofEngland’s The Big Breakfast, to Canada’s Naked News, during which anchorspresent a newscast while doing a striptease.40


39. Fallows (1996).40. Lynn Hirschberg, “Giving Them What They Want,” New York Times Magazine, September

4, 2005.


These solutions—comedy and sex—are only two of many possible alternativeattention-generating mechanisms. But those pursuing the solution need to moveaway from the premise that commercial pressures alone are responsible and real-ize that this is a problem that has been faced before, albeit in very different contexts.Children’s television, for example, was once widely bemoaned as antisocial in itsconsequences. Kids were particularly drawn to highly arousing, violent programs—precisely the kind of content most feared by parents. Yet through collaborationswith researchers, organizations such as the Children’s Television Workshop wereable to draw on knowledge of human behavior to come up with alternative meansof grabbing children’s attention. The early evidence on entertainment-orientednews programs is promising in demonstrating that audiences do learn somethingabout politics from their content.41 While it is not unreasonable to be optimisticthat answers will be found, it seems foolhardy at this point to try to predict thefuture of political media.

The underlying question that still needs to be confronted—by scholars as wellas those in the media business—is how to make a topic that is not inherentlyinteresting to many Americans nonetheless exciting to watch. And if the answeris not behind-the-scenes coverage of election strategy, or mudslinging on politicaltalk shows, or partisan extremists rallying the troops, then what will keep thosepolitically marginal citizens from watching movies on cable instead?

248 diana c. mutz

41. Baum (2003).


how the mass media divide us - university of pennsylvania

Documents